Viewing Inventory Normalization Insights

Important:Previously, IT Visibility provided an Inventory Normalization Statistics report (available at Data Collection > IT Visibility Inventory Status > Inventory Normalization Statistics). With Technology Intelligence Platform, this report is replaced by the Inventory Normalization Insights report (available at Data Collection > IT Visibility Inventory Status > Inventory Normalization Insights).

Inventory normalization insights give your greater understanding as to how hardware and software inventory has been normalized.

To view inventory normalization insights:

1.

Navigating to the Inventory Normalization Insights page (Data Collection > IT Visibility Inventory Status > Inventory Normalization Insights).

The Inventory Normalization Insights page appears.

2.

Click the Hardware or Software tab and review the following subsections for descriptions of these statistics:

•

Hardware Normalization Statistics

•

Software Normalization Statistics

Hardware Normalization Statistics

Hardware Normalization Statistics

Statistic	Description
Per discovery source	Evidence count is the count of evidence (that is, the information collected from a system to detect the existence of the hardware) from each discovery source being ingested in the normalization process.
Total evidence count	This statistic is broken down by evidence type: • Consolidated evidence indicates the evidence that is retained as a result of merging, consolidation or deduplication process amongst multiple evidence, which are used to recognize the exact same hardware. • Duplicate evidence indicates the evidence that is not retained as a result of merging, consolidation or deduplication process because it is considered as a duplicate amongst multiple evidence, which are used to recognize the exact same hardware.
Consolidated inventory count	This statistic is broken down by recognition status: • Recognized to model indicates that, out-of-the-box, the evidence is fully recognized (that is, associated with a Technopedia catalog entry) all the way to the hardware model level. • Recognized to product indicates that, out-of-the-box, the evidence is only partially recognized to the hardware product level. • Recognized to manufacturer indicates that, out-of-the-box, the evidence is only partially recognized to the manufacturer level. • Unrecognized indicates that, out-of-the-box, the evidence is not recognized by the normalization process. Therefore, the evidence will go through the gap-fill process.
Inventory count	This statistic is broken down for reach recognition status by: • Active is the count of hardware that is managed and taken into consideration when calculating the final inventory count for further downstream use-cases (such as license consumption calculation). • Ignored is the count of hardware that is not taken into consideration when calculating the final inventory count for various reasons (for example, the hardware is being retired, disposed of, etc.).

Software Normalization Statistics

Software Normalization Statistics

Statistic	Description
Per discovery source	Evidence count is the count of evidence (that is, the information collected from a system to detect the existence of the software) from each discovery source being ingested in the normalization process.
Total evidence count	This statistic is broken down by evidence type: • Installer evidence indicates the evidence collected from a local (device-based) list of installed software, such as Add/Remove Programs or MSI. • File evidence indicates evidence in the form of a file that is found on a computer—it may simply be the name of the file, or the information found within the file (such as with ISO-compliant software identification [SWID] tags). Note the following: • While software recognized by file evidence is available in IT Visibility, statistics on file evidence are not yet available. • To establish accurate software inventory recognition, IT Visibility retrieves mappings from both the Application Recognition Library (ARL) and the Technopedia library. Non-executable (for example, .sys, .swtag, .swidtag files) mappings primarily come from ARL, while executable (for example, .exe files) mappings come from both ARL and Technopedia. • Other evidence indicates evidence other than installer or file evidence. This currently includes all evidence used to recognize operating systems, including OS-related WMI evidence, OS evidence from third-party discovery systems, and OS evidence from commands such as os-release on Linux, sw_vers on MacOS, and uname on other Unix systems.
Installer Evidence count	This statistic is broken down by recognition status: • Recognized indicates that, out-of-the-box, the evidence is recognized (that is, associated with a Technopedia catalog entry) to either the software release level or the software product level. • Unrecognized indicates that, out-of-the-box, the evidence is not recognized by the normalization process. Therefore, the evidence will go through the gap-fill process.
Recognized evidence count	This statistic is broken down for reach recognition status by: • Consolidated evidence indicates that the evidence is retained as a result of merging, consolidation or deduplication process amongst multiple evidence, which are used to recognize the exact same software. • Duplicate evidence indicates that the evidence is not retained as a result of merging, consolidation or deduplication process because it is considered as a duplicate amongst multiple evidence, which are used to recognize the exact same software. • Irrelevant evidence indicates that the evidence has been recognized as not having any value to the customer (for example, it’s a device firmware, a driver, a miscellaneous application add-on, plugin or other negligible files, etc.). Therefore, the evidence is separated out from the calculation of the software install count.
Consolidated evidence	This statistic identifies software that is installed in the inventory and is broken down by: • Active is the count of software that is detected on an active hardware. Therefore this evidence is taken into consideration for further downstream use-cases (such as license consumption calculations). • Ignored is the count of software that is detected on an ignored hardware.